Meteorology   

possibly have springs. The waters we find on the earth either flow

or are stationary. All flowing water has springs. (By a spring, as

we have explained above, we must not understand a source from which

waters are ladled as it were from a vessel, but a first point at which

the water which is continually forming and percolating gathers.)

Stationary water is either that which has collected and has been

left standing, marshy pools, for instance, and lakes, which differ

merely in size, or else it comes from springs. In this case it is

always artificial, I mean as in the case of wells, otherwise the

spring would have to be above the outlet. Hence the water from

fountains and rivers flows of itself, whereas wells need to be

worked artificially. All the waters that exist belong to one or

other of these classes.

On the basis of this division we can sec that the sea cannot have

springs. For it falls under neither of the two classes; it does not

flow and it is not artificial; whereas all water from springs must

belong to one or other of them. Natural standing water from springs is

never found on such a large scale.

Again, there are several seas that have no communication with one

another at all. The Red Sea, for instance, communicates but slightly

with the ocean outside the straits, and the Hyrcanian and Caspian seas

are distinct from this ocean and people dwell all round them. Hence,

if these seas had had any springs anywhere they must have been

discovered.

It is true that in straits, where the land on either side

contracts an open sea into a small space, the sea appears to flow. But

this is because it is swinging to and fro. In the open sea this motion

is not observed, but where the land narrows and contracts the sea

the motion that was imperceptible in the open necessarily strikes

the attention.

The whole of the Mediterranean does actually flow. The direction

of this flow is determined by the depth of the basins and by the